Model-free estimation of outdoor performance of a model epoxy coating system using accelerated test laboratory data

Abstract

Laboratory and field exposure results for thin transparent films of a model epoxy coating system have been successfully linked using both dosage and dose, assuming only that the additivity law and the reciprocity law are valid. This paper describes the technique used in the linkage step. Laboratory exposures were carried out on the National Institute of Standards and Technology (NIST) Simulated Photodegradation by High Energy Radiant Exposures (SPHERE) using a factorial design of four temperatures, four relative humidities, four ultraviolet-visible ranges, and four neutral density filters. Similar specimens were exposed on the roof of a NIST laboratory in Gaithersburg, MD. The temperature and relative humidity of the outdoor exposures and the solar spectrum were used to characterize the roof environment at 12 min intervals. Chemical degradation of all specimens was followed by transmission Fourier Transform Infrared (FTIR) spectra. Dose, the radiation incident on the specimens, was estimated from lamp and filter spectra for the SPHERE exposures and from the solar spectra for outdoor exposures. Specimen UV–visible absorbance spectra were used with the estimates of dose to give estimates of dosage—the energy absorbed by the specimens. Using filter ranges for the SPHERE data allowed estimation of relative degradation efficiency over the range of the solar spectrum. This was used implicitly by the computer program to estimate the degradation in specimens exposed outdoors. The paper describes a model-free heuristic approach that automatically predicts chemical degradation outdoors using the outdoor environment parameters and the SPHERE laboratory data. That both dose and dosage are equally valid methods of successfully predicting outdoor degradation may be a consequence of all the films being approximately the same thickness (about 6 μm).